In　this　paper,　we　propose　a　novel　method　based　on　color　space　volume　and　multimodal　histogram　thresholding　(MHT)　for　leukocyte　segmentation　in　human　blood　smear　images　under　the　rapid　staining　condition.　The　proposed　method　first　defines　a　new　color　space　volume　to　highlight　the　leukocyte＇s　nucleus　region,　and　it　then　uses　the　MHT　method　to　extract　the　nucleus.　Then,　the　proposed　method　takes　the　central　region　in　the　superpixel　segmentation　result　as　an　initial　leukocyte　region,　and　it　then　uses　the　MHT　method　to　extract　the　leukocyte　region.　Finally,　the　cytoplasm　is　obtained　by　subtracting　the　nucleus　region　from　the　leukocyte　region.　Experimental　results　on　a　dataset　containing　300　leukocyte　images　under　rapid　staining　condition　show　that　the　proposed　method　provides　a　higher　segmentation　accuracy　than　the　counterparts.　©　2019　IEEE.